Technical Account Manager

Ambient.ai introduces new agentic physical security capabilities. Ambient.ai has announced new capabilities across the Ambient Platform, adding agentic intelligence to how security teams monitor, investigate and operate. The new capabilities include an AI agent, a new case management workflow and infrastructure upgrades that double camera density on existing hardware. "Recording video was never the hard part," said Vikesh Khanna, CTO & Co-founder, Ambient.ai. "The hard part is understanding what's happening on a site when it actually matters. "Among the set of new capabilities, Agentic Video Walls give operators an agent that monitors every camera on their behalf, and Case Management, an automated workflow that, paired with Semantic Search, turns hours of manual scrubbing into a fully connected investigation in minutes. "That's the difference between a simple system of record and a platform that reasons alongside your team." The Ambient Platform introduces enhancements across three core areas, situational awareness, investigations and operational scale. Continuous situational awareness. Agentic Video Walls - Brings an AI agent to every connected camera on site, acting like a relentless, hyper-focused editor: scanning every stream, deciding what belongs on the front page and surfacing it in real time. The agent monitors around the clock and surfaces the single most relevant event once every 60 seconds, paired with an AI-generated, plain-language description of what's happening. Rapid investigations and response. Case Management - Turns scattered manual work into one continuous flow. After tracking a person across cameras using natural language through Semantic Search, operators record clips directly into a case as they're found, then generate a fully editable, AI-written narrative with one click, complete with auto-sorted clips and metadata. From there, cases can be downloaded as a compressed, sequenced folder or shared through expiring, revocable links with a full audit trail. Semantic Search Streamlining - Semantic Search is the platform's engine that lets operators describe who or what they're looking for in plain language, feeding straight into a case instead of hours of manual scrubbing. This release sharpens results across the board: matching frames of the same person or vehicle across a five-minute span now group into a single result and results filter down to a specific camera or group, Ambient says. Enterprise scale and cost efficiency. Double Stream Density - Ambient Foundation now supports up to 2x the camera streams on a single Ambient Edge Appliance, made possible by a smaller, more focused application that frees up compute and memory per stream. Existing customers get double the streams on current appliances, and new sites need half as many appliances to cover the same camera count.

Ambient.ai: PACS alarm management solutions. 21 May 2026 Quick Read Summary is AI-generated, newsdesk-reviewed * Ambient.ai introduces major updates to its Ambient Access Intelligence platform for better PACS management. * New capabilities address chronic issues, enhance door security, and reduce alarm noise by 95%. * Integrates with major PACS systems, reducing costs and saving time with AI-driven solutions. Ambient.ai has unveiled significant enhancements to Ambient Access Intelligence, its advanced module for managing physical access control systems (PACS). These expansions incorporate two new features: 'Doors with Issues' and 'Doors Unsecured'. These improvements provide users with tools to diagnose persistent infrastructure problems and monitor physically unsecured doors in real-time, offering a comprehensive view for PACS alarm management. This development marks a substantial shift in access control operations, extending its capabilities beyond efficient alarm handling into a proactive model that seeks to prevent unnecessary incidents and focus resources effectively. Streamlining alarm responses. Traditional access control systems can inundate security personnel with overwhelming alarm signals, sometimes prompting operators either to spend excessive time addressing them or to ignore them, potentially increasing security risks. However, Ambient Access Intelligence is designed to mitigate this issue by decreasing PACS alarm volume by 95% through pre-emptive resolution of false positives. By doing so, the system ensures that operators concentrate on validated events only. Traditional access control systems can inundate security personnel with overwhelming alarm signals Vikesh Khanna, CTO and Co-Founder of Ambient.ai, explains, "Alarms without intelligence are just noise. Our Reasoning AI platform has already turned millions of PACS alarms into verified signals for our customers. With the addition of 'Doors with Issues' and 'Doors Unsecured', Ambient Access Intelligence is now entering its next phase: not only clearing alarms, but deciding which events deserve a response, preventing the issues that generate noise in the first place, and verifying that every door is physically secure." Advancements in AI-driven analysis. The Ambient Access Intelligence platform comprises three core components that work cohesively to enhance security operations: Alarm auto clearing. This capability integrates real-time video with every PACS alarm via the PACS Correlation Engine to automatically resolve frequent false positives, such as exit events, interaction with door interiors, wind interference, sensor faults, or self-restoring doors. As a result, over 95% of alarms are resolved without human intervention, with only validated incidents being presented as detailed alerts. Doors with Issues. This new feature offers security administrators an AI-driven analysis to pinpoint chronic noise sources due to infrastructure issues. It organises issues into categories like Door & Alarm Behaviour, Camera Health, and Camera-Reader Mapping, complete with identified root causes and visual evidence. This facilitates seamless communication with facilities and maintenance teams, supplanting slow manual assessments. Doors Unsecured. Providing real-time operational insight, this feature ensures security operators can verify physical door closures by confirming door-restored PACS signals. A live video feed of unsecured doors spans all premises, aiding coordinated security measures and eliminating isolated incident responses. Typically, access-controlled doors spend around 14.5 hours unsecured weekly, a gap now visible for prompt action. Creating a self-validating security perimeter. These capabilities collectively transition access control from a reactive model into a self-validating perimeter, enabling the efficient use of human resources on critical events. At ServiceNow, the adoption of Ambient Access Intelligence has led to the processing of over 240,000 alarms, auto-clearing 94% and realising savings of more than 15,000 labour hours and over $500,000 in avoided costs. The platform integrates seamlessly with several leading PACS systems, including LenelS2, Honeywell Pro-Watch, Genetec, Milestone, Allegion, Software House C-CURE 9000, Avigilon, Brivo, Kastle, Genea, among others. Ambient Access Intelligence requires only the Ambient Foundation as its base layer, facilitating deployment on existing infrastructure without necessitating extensive replacements.

Ambient.ai, a physical security AI startup, has doubled its new annual recurring revenue in FY26 compared to FY25, with net revenue retention exceeding 140% as its customer base tripled. Multiple Fortune 100 clients have expanded to seven-figure contracts. The Redwood City-based company processes over 200 million video hours and 10 billion events daily through its platform. It recently launched Ambient Pulsar, the first edge-optimised reasoning Vision-Language Model for physical security, trained on over one million hours of enterprise video. Ambient.ai's clients include Cisco, ServiceNow, SentinelOne and TikTok. ServiceNow's deployment achieved a 94% false alarm reduction and saved 15,069 hours of manual triage. The company maintains a sub-1x burn multiple whilst delivering frontier-model reasoning performance at up to 50 times higher efficiency than general-purpose AI models.